Dramatically different phenotypes in mouse models of human Tay-Sachs and Sandhoff diseases
Dramatically different phenotypes in mouse models of human Tay-Sachs and Sandhoff diseasesDaniel Phaneuf1,+, Nobuaki Wakamatsu1,, Jing-Qi Huang1, Anita Borowski5, Alan C. Peterson2, Sheila R. Fortunato6, Gerd Ritter6, Suleiman A. Igdoura3, Carlos R. Morales3, Guylaine Benoit1,4, Beverly R. Akerman1, Daniel Leclerc1, Nobuo Hanai7, Jamey D. Marth5,, Jacquetta M. Trasler1,4 and Roy A. Gravel1,*
1Research Institute, Montreal Children's Hospital, Departments of Pediatrics and Human Genetics, McGill University, 2The Royal Victoria Hospital Research Institute, Department of Neurology and Neurosurgery, McGill University, 3Department of Anatomy and Cell Biology, McGill University, 4Department of Pharmacology and Therapeutics, McGill University, Montreal, Canada, 5Biomedical Research Centre, University of British Columbia, Vancouver, Canada, 6Ludwig Institute for Cancer Research, New York Unit at the Memorial Sloan-Kettering Cancer Center, New York, USA and 7Division of Immunology, Tokyo Research Laboratories, Kyowa Hakko Kogyo Co. Ltd, Tokyo, Japan
Received October 17, 1995;Revised and Accepted October 27, 1995
We have generated mouse models of human Tay-Sachs and Sandhoff diseases by targeted disruption of the Hexa ([alpha] subunit) or Hexb ([beta] subunit) genes, respectively, encoding lysosomal [beta]-hexosaminidase A (structure, [alpha] ) and B (structure, [beta][beta]). Both mutant mice accumulate GM2 ganglioside in brain, much more so in Hexb -/- mice, and the latter also accumulate glycolipid GA2. Hexa -/- mice suffer no obvious behavioral or neurological deficit, while Hexb -/- mice develop a fatal neurodegenerative disease, with spasticity, muscle weakness, rigidity, tremor and ataxia. The Hexb -/- but not the Hexa -/- mice have massive depletion of spinal cord axons as an apparent consequence of neuronal storage of GM2. We propose that Hexa -/- mice escape disease through partial catabolism of accumulated GM2 via GA2 (asialo-GM2) through the combined action of sialidase and [beta]-hexosaminidase B.
Gangliosides are glycosphingolipid constituents of the outer surface of most cell membranes. They are particularly abundant in the nervous system (1 -3 ) where they may function in neuritic growth, cell recognition, and neural transmission (4 ,5 ). They are catabolized in the lysosome where glycohydrolases degrade them by sequential removal of terminal sugars to a ceramide core (see Fig. 1 for nomenclature). Inherited defects of the hydrolytic enzymes cause lysosomal storage diseases in which glycolipid intermediates accumulate in the lysosome.
Targeting vectors were developed from previously characterized clones containing the mouse Hexa and Hexb genes (17 ,18 ). The construct used for targeted disruption of the Hexa gene was interrupted in exon 11 by a neomycin-resistance gene cassette (Fig. 2 A). Nine of 153 independent CGR8 embryonic stem (ES) cell clones were identified in which one Hexa allele had incorporated the desired mutation (Fig. 2 B). Two of three clones injected into C57BL/6J blastocysts gave rise to 21 chimeric animals with ES cell contributions ranging from 5 to 95% as judged by the proportion of agouti coat color. One of eight chimeric males bred to C57BL/6J wild-type females transmitted the disrupted Hexa allele through the germline. Intercrosses between heterozygous mice produced progeny with no apparent reduction in the number of pups per litter. PCR analysis of tail biopsies (Fig. 2 C) used to determine the genotype of 215 offspring of heterozygous matings gave 55 wild-type (25%), 117 heterozygous (53%) and 50 homozygous (22%) offspring at the expected Mendelian ratio (1:2:1). Hexa -/- mice showed no apparent difference in size, behavior, or reproductive ability compared with wild-type litter mates up to 1 year (Fig. 3 A).
The Hexb construct contained a neomycin-resistance gene cassette inserted in exon 2 (Fig. 2 D). A total of 3/90 CGR8 and 3/110 R1 independent ES clones were confirmed to have a correct gene targeting event (Fig. 2 E). Twenty-four chimeric mice were obtained with ES cell contributions ranging from 10 to 95% as judged by the proportion of agouti coat color. Nine chimeric males were bred to C57BL/6J wild-type female and five of them (two from CGR8 and three from R1) transmitted the disrupted Hexb allele through the germline. Heterozygous Hexb +/- mice were phenotypically normal and litters resulting from intercrosses between heterozygotes were normal in size. Using PCR (Fig. 2 F), genotypic analysis of tail biopsies from 140 offspring mice showed that homozygous Hexb (-/-) mice were born at a frequency of 23%, indicating the absence of prenatal homozygous lethality.
Twenty-seven homozygous Hexb (-/-) mice (10 male, 17 female), derived from both CGR8 and two of the R1 lines, appeared healthy at birth and were indistinguishable from their normal litter mates in behavior and reproductive ability up to 3-4 months of age. However, from 111 to 147 days after birth, the affected mice became lethargic and began to show tremor, muscle weakness, stiffening of the hind limbs and an ataxic gait (Fig. 3 B-D). Movement was slow and they had a tendency to drag the hind limbs. They were unable to right themselves when placed on their back. Weight loss became apparent, possibly due to an inability to feed (dysphagia). Preterminal symptoms included spastic quadriparesis, tremor, and myoclonus (both startle and non-startle). The disease was inevitably fatal in homozygous mice within 6 weeks of onset.
Messenger RNA levels were examined in tissues from wild-type, heterozygous and homozygous mice by Northern blot using specific cDNA probes (Fig. 4 ). A ~2.2 kb Hex [alpha] transcript was detected in all tissues analyzed from wild-type and Hexa +/- mice, but was absent in Hexa -/- mice. In Hexa +/- and -/- mice, an RNA species of ~3.2 kb was observed (Fig. 4 B,C). It was also detected with a neo probe (data not shown) and appeared to be due to a Hexa-neo fusion, presumably derived from the inclusion of all or a portion of the neomycin cassette by aberrant splicing. Hexb mRNA was undetectable in most tissues analyzed from Hexb -/- mice. An exception was testis in which a shorter transcript of 1.8 kb was observed (Fig. 4 E,F). As the neo sequence was not detected in the shorter species (data not shown), it was likely produced by exon skipping. While a very faint 1.8 kb band could also be detected in brain and epididymis (Fig. 4 F), no enzymatically active product resulted from the transcript (see below).
Brain and liver extracts of wild-type, Hexa -/- and Hexb -/- mice were assayed for activity toward the synthetic substrates, 4-methylumbelliferyl-[beta]-D-N-acetylglucosamine (MUG), which is hydrolyzed by all Hex isoenzymes, and 4-MUG-6-sulfate (MUGS), which is hydrolyzed only by Hex isoenzymes containing an [alpha] subunit (7 ). Mouse Hex A and Hex B, with similar subunit compositions as the human enzyme, behave similarly toward MUG and MUGS (19 ). Both extracts from Hexa -/- mice contained approximately 50% of normal MUG activity, while MUGS activity was reduced to less than 5% in brain and less than 1% in liver (data not shown). In the Hexb -/- mice, the activity with either substrate was reduced to less than 1.5%. These data suggested that, in the Hexa -/- mouse, Hex B activity was retained while Hex A activity was reduced to near background values. In order to establish the isoenzyme distribution of the observed activity, we used chromatofocusing (20 ). In the brain homogenate from a wild-type mouse (Fig. 5 , wild type), the first MUG-active enzyme to elute from the column was Hex B with a peak centered at pH 6.9 (range 7.0-6.5), followed by Hex A with a peak centered at pH 4.8 (range 5.1-4.5). Hex activity found in the region ranging from pH 5.8 to 6.1 of the elution profile may correspond to incompletely processed `intermediate' pI forms of Hex (Hex I) (21 ,22 ). Analysis of the brain extract from a Hexa -/- mouse showed abundant Hex B isoenzyme eluting in the expected fractions and undetectable enzyme activity in the Hex A fractions (Fig. 5 , Hexa -/-). Brain extracts from a Hexb -/- mouse revealed negligible activity in the Hex B and Hex A fractions. Minimal activity of a third form of Hex, Hex S, an [alpha] dimer found in human Sandhoff disease (23 ), was recovered by step elution with sodium citrate buffer, pH 3.5 (Fig. 5 Hexb -/-). There was also trace activity eluting at pH 4.4-4.7 that may correspond to Hex S precursor (20 ). A similar chromatofocusing done with testis extracts from Hexb -/- mice had background activity, confirming that the abundant 1.8 kb [beta]-subunit transcript did not encode a functional protein (data not shown).
Glycosphingolipids obtained from brain and liver were examined by HPTLC after separation into neutral and acid fractions by DEAE-Sephadex A25 chromatography. GM2 and GA2 were specifically identified and quantitated by immuno-thin layer chromatography (ITLC) using the mAb KM966 to detect GM2 ganglioside and mAb 2D4 for detection of GA2. The HPTLC revealed the brains of wild-type mice to contain predominantly GM1, GD1a, GD1b, GT1b and ceramide monohexosides (data not shown). In brains from Hexb -/- mice, glycolipids with the mobilities of GM2 and GA2 were detected in high abundance, while in Hexa -/- mice, only GM2 was elevated (Fig. 6 A,B). Quantitation by ITLC showed that Hexb -/- mice brains contained GM2 over 10-20* elevated and GA2 8-50* elevated (Fig. 6 C). In contrast, in the Hexa -/- mice, brain GM2 was elevated only 6-10* control values and GA2 was within normal range.
In the liver of normal mice, GM2 was the major ganglioside, while GA2 was very low or below the limit of detection. No significant change in the level of GM2 was observed in Hexa -/- (unchanged) or Hexb -/- (1-3* elevation; Fig. 6 C). However, Hexb -/- but not Hexa -/- mice contained GA2 about 10-30* the level detected in normal mice. Thus, the Hexa -/- mice accumulated neither of these lipids in liver.
Macroscopic examination of the brain of Hexa -/- and Hexb -/- mice revealed a normal appearance. However, light microscopy of semi-thin sections (0.5 [mu]m thick) of cerebral cortex revealed neurons with clusters of perinuclear inclusions in both the Hexa -/- and Hexb -/- mice. Ultrastructurally, the inclusions were similar to MCB observed in the human diseases. They varied from concentric to parallel membranous arrays to crescent-shaped inclusions and amorphous structures (Fig. 7 A,B).
The distribution of GM2 ganglioside was determined by immunohistochemistry using mAb KM966. In normal brain, GM2 was undetectable, except in rare cells. In Hexa -/- brains, GM2 showed a specific and limited distribution, although staining could be very intense in positive neurons. Coronal sections of Hexa -/- brain showed GM2 immunostaining in the septal area, hippocampus and the pyramidal layer of the cerebral cortex. Intense staining was observed in a thin band of cells, layer 6b, at the base of the cortex (Fig. 9 A). Staining was observed in scattered cells of the basal ganglia, and the thalamus and hypothalamus showed sparse to heavy staining in different regions with intense staining of the lateral mammillary nucleus (Fig. 9 B). The cerebellum was unstained (Fig. 9 C) except at the tip of one lobule (the simple lobule, not shown). The brainstem showed a single site of intense staining in the large cells of the mesencephalic nucleus of the 5th nerve. The cell bodies of the spinal cord, including anterior horn cells, did not show GM2 ganglioside immunostaining (Fig. 9 D).
The mutant mice we have produced appear to be valid models of GM2 gangliosidosis. Hexa -/- mice lack Hex A activity and Hexb -/- mice lack both Hex A and Hex B activity. Both accumulate GM2 ganglioside and have characteristic lysosomal inclusions in neuronal cells. However, unlike the human disease, Hexa -/- mice have no behavioral phenotype, while Hexb -/- mice express a rapidly progressive, fatal neurodegenerative disease. Our observations on Hexa -/- mice are similar to those reported by Proia and colleagues who also produced a Hexa -/- mouse by gene targeting (25 ,26 ). While the mutant we generated was disrupted by an insertion in exon 11, theirs was disrupted by an insertion in exon 8. The phenotype expressed by the two independently derived Hexa -/- mice confirms the benign nature of a complete deficiency of Hex A activity in mice. The similarity of the GM2 ganglioside distribution in the brains of the two Hexa -/- mice and the much greater accumulation in Hexb -/- mice suggest that the fundamentals of ganglioside biology are similar in mice and humans. Consequently, the disease avoidance in Hexa -/- mice must be due to the single biochemical difference between the two mice, the retention of Hex B activity in the Hexa -/- mouse.
The striking progression of the neurodegenerative process in our two independently derived (CGR8 and R1) lines of Hexb -/- mice parallels the rapid downhill course of human Tay-Sachs or Sandhoff disease (6 ). In both, the newborn appears healthy. Muscle weakness and lack of purposeful movement appear as early symptoms and both progress to an immobile, unresponsive state. Tremor, ataxia, spasticity and muscle weakness are features of later onset, more slowly progressive forms of the human disease that we observed in Hexb -/- mice. The absence of GM2 storage in the brain of Hexb -/- mice in the first weeks after birth readily explains the absence of early symptoms. In humans, the first appearance of MCB in the anterior horn cells of the spinal cord was in a fetus with Tay-Sachs disease at 12 weeks of gestation (27 ). In a similar manner, in a Hexb -/- mouse, GM2 storage was first detected in anterior horn cells observed 2-3 weeks post-natally.
The remarkable extent of axon damage, the extensive vacuolation of neurons in the gray matter including motor neurons, and the large, MCB-engorged axon remnants in the spinal cord are sufficient to explain the fatal nature of the Hexb -/- disease. These pathological findings suggest that the integrated circuitry of the spinal cord is disrupted at multiple levels, leading to total incoordination of somatic motor systems. While the extensive storage of GM2 ganglioside in the basal ganglia, cerebellum and cerebral cortex may also have contributed to the observed symptoms, we do not yet know the extent to which intact neurons, swollen with GM2 storage bodies, are able to function.
The quantitative analysis of brain and liver glycosphingolipids provides support for a threshold model of GM2 toxicity. First, there was 1.5-3 times the level of GM2 in the brains of a Hexb -/- compared with Hexa -/- mice. These results were consistent with the relatively low intensity of immunostaining in the brain of Hexa -/- mice. Further, as GA2 is derived by catabolism of GM2, the total storage is more aptly represented by GM2 + GA2. It equalled 440-1452 nmol/g wet weight of Hexb -/- brain, in striking contrast to the 194 and 234 nmol/g wet weight total storage obtained in the two Hexa -/- brains. The critical, toxic threshold may reside between these two ranges, as it is not known whether GA2 also contributes to neuron dysfunction. Accumulated GM2 has been shown to induce ectopic sprouting of neurites in cortical pyramidal cells (28 ) and may also affect neurotransmitter function (29 ).
A plausible hypothesis to explain disease progression is that GM2 slowly accumulates in neurons without a detectable influence on behavior in the affected animal. Once the critical threshold is achieved, abnormalities of neuronal structure and function are induced that result in subsequent disintegration of axons and, perhaps, cell bodies in spinal cord and presumably in brain. As the severity of the nerve tract depletion increases, the affected mice begin to show an overt phenotype which manifests itself as motor dysfunction. After 3-4 months of unremarkable behavior, the mice enter a rapid phase that leads to complete immobilization within 6 weeks of the first appearance of symptoms.
Our combined data suggest that the Hexa -/- mice escape disease by keeping the GM2 stores to below toxic levels. The presence of the Hex B isoenzyme offers a compelling clue to a mechanism for metabolic sparing in these mice. Two models for partial metabolism of GM2 are envisioned (Fig. 1 ). In Model 1, Hex B acts directly on GM2 to convert it to GM3, which is further catabolized by sialidase to Lac-Cer. This model would readily explain the absence of symptomatic disease in Hexa -/- mice. However, it would also imply that the substrate specificity of Hex B was different in mice and humans, as, in the latter, Hex B shows no activity toward GM2 (30 ). In early studies on the GM2 activator, Burgh et al. (31 ) demonstrated that mouse brain extracts from which Hex A was removed could release GalNac from GM2, suggesting to them that Hex B in combination with the GM2 activator could act on GM2. These experiments were done using 3H-GM2 labelled in the GalNAc moiety so that it was not possible to distinguish the direct action of Hex B from a combined action of sialidase followed by Hex B. Further, mouse and human Hex B behave similarly toward synthetic substrates that distinguish Hex A and Hex B activity (19 ), as confirmed in Figure 5 . The trace activity of MUGS toward Hex B in Figure 5 is consistent with a high Km for MUGS, as determined for the human enzyme (7 ). Thus, it is unlikely that a sparing effect of Hex B could occur through direct action of the enzyme on GM2.
According to Model 2 (Fig. 1 ), sialidase acts on GM2 to convert it to GA2, which is further degraded by Hex B to Lac-Cer. The activity of sialidase toward GM2 is not normally favored owing to the more potent activity of Hex A toward GM2 (6 ). However, in the absence of Hex A due to mutation in the Hexa -/- mouse, partial catabolism via GA2 appears to be prominent. Although lysosomal sialidase is only weakly active toward GM2 (32 ,33 ) and Hex B hydrolyses GalNAc from GA2 at only ~3% the activity of Hex A (30 ), there is nonetheless a several-fold higher level of GA2 in autopsy brains of Sandhoff vs Tay-Sachs patients (34 ). We surmise that the difference in the Hexa -/- mouse that allows it to escape disease is a more efficient conversion of GM2 to GA2 than occurs in the human disease. An example of a functioning sialidase-mediated GM2 catabolic pathway is found in the mouse neuroblastoma line Neuro2a. Riboni et al. (35 ) treated Neuro2A cells with 3H-GM2 labelled in the sphingosine moiety and showed that the principal metabolites were GA2 and Lac-Cer. A similar incubation with 3H-GM1 also resulted in catabolism through GA2.
Our data favor Model 2. In the absence of Hex B in Hexb -/- mice, stored GA2 is derived directly from GM2. In the Hexa -/- mouse, the same conversion of GM2 to GA2 occurs but, in this instance, the GA2 is more rapidly cleared due to the presence of Hex B, driving the reaction toward further catabolism.
At the completion of this investigation, Proia and colleagues (36 ) reported a Hexb -/- mouse generated by targeted disruption of exon 13 of the Hexb gene. They describe a similar behavioral phenotype with onset at about 3 months after birth. Glycolipid storage was demonstrated with periodic-acid-Schiff (PAS) in almost all neurons of the central nervous system. They ascribed the motor deficit in their mice to the prominence of storage neurons in areas such as the cerebellum and spinal cord. It is not known if their mice also demonstate the marked depletion of axons of the spinal cord, which we believe is more directly compatible with the very rapid course of the disease. They went on to examine 3H-GM1 catabolism in embryonic fibroblasts cultured from Hexa -/- and Hexb -/- mice and demonstrated the dominance of the GA2 pathway (Model 2), even in normal cells. They concluded that the mouse sialidase has a much greater affinity for GM2 than the human enzyme, making the capacity to escape disease in the Hexa -/- mouse a unique property of this mouse model. It remains to be seen whether the GA2 bypass occurs due to an intrinsic property of sialidase, as they suggest, or to an increased level of sialidase activity.
The results of this investigation highlight the pharmacologic modulation of brain sialidase levels as a potential therapeutic strategy for human Tay-Sachs disease. Specifically, inducing sialidase to higher specific activity could result in the more efficient conversion of GM2 to GA2 and consequent continued catabolism via Hex B. If this is the mechanism conferring the benign phenotype in the mouse, it seems possible that infants treated shortly after birth could also maintain GM2 at subtoxic levels and escape symptomatic disease.
To construct the targeting vector, pHA12neo (Fig. 2 A), a 1.7 kb SmaI-HindIII fragment which included exons 11-13 of the mouse Hexa gene (17 ) was subcloned in pBluescript (Stratagene). A 1.8 kb EcoRI-XbaI fragment containing the neomycin-resistance (neor) gene controlled by the phosphoglycerate kinase-1 (pgk) promoter (PGKneobpA; 37 ), kindly provided by Dr J. Rossant, Toronto, Canada, was blunted with Klenow and inserted into the unique EcoRV site in exon 11, disrupting the coding sequence at base pair 1216 (corresponding to the Hexa cDNA; 38 ). The neor cassette was subcloned in the same transcriptional orientation as the genomic sequences. A 5.2 kb HindIII-SalI fragment containing exon 14 and the 4.8 kb BstI-SmaI fragment containing exons 6-10 were subcloned upstream and downstream of exons 11-13, respectively.
To construct the targeting vector, pHexb11.3neo (Fig. 2 D), a 5.6 kb ClaI-KpnI fragment containing exons 1 and 2 of the mouse Hexb gene (18 ) was subcloned into pBluescript. The 1.8 kb neor expression cassette was inserted into a blunted, unique XhoI site in exon 2 disrupting the coding sequence at base pair 346 (corresponding to the Hexb cDNA; 39 ). Finally, a 5.7 kb KpnI-SmaI fragment which contains exons 3-5 of the mouse Hexb gene was included at the 3' end of the construct.
The CGR8 ES line, kindly provided by Dr William C. Skarnes, Edinburgh, UK), and the R1 ES line (40 ) were grown on 0.1% gelatin-coated culture dishes in medium as described, except that the R1 line included the use of irradiated early passage mouse primary embryonic fibroblasts (feeders) (40 ). Twenty [mu]g of the targeting constructs, linearized with Not1, were used to transfect 2*107 ES cells in 800 [mu]l of cold PBS using a Bio-Rad GenePulser (240V, 500 [mu]F). After the electroporations, cells were seeded at varying dilutions in ES culture medium and G418 (180 [mu]g/ml; Geneticin, GIBCO) was applied 24 h after transfection. After 7-10 days, G418-resistant clones were isolated, resuspended and plated into two wells, one for freezing down and the other for DNA analysis (41 ). Genomic DNA (10 [mu]g) from individual ES clones was digested to completion with BamH1 (Hexa knockout) or EcoRV or KpnI (Hexb knockout) and analyzed by Southern blot. The hybridization cDNA probes were a 156 bp 5'-flanking fragment (Hexa knockout; Fig. 2 , probe `a') or a 202 bp 3'-flanking fragment (Hexb knockout; Fig. 2 , probe `c') containing sequences not included in the targeting vectors. Clones identified as being heterozygous for the desired homologous recombination event were further analyzed by Southern blot using internal cDNA probes (Fig. 2 ; probes `b' and `d').
The generation of chimeric mice from Hexa or Hexb targeted ES cells were performed as previously described (42 ). Newborn pups were identified as chimeric on the basis of agouti coat color. Highly chimeric males were crossed at 6 weeks of age with C57BL/6J females. Germline transmission was scored on the basis of agouti coat color of F1 pups. Heterozygotes were interbred to produce homozygous offspring. Mutant gene transmission was confirmed by Southern blot analysis of mouse tail DNA.
PCR typing of tail DNA was performed as shown schematically in Figure 2 . Primers for detection of Hexa sequences were: primer 1, 5'-GGCCAGATACAATCATACAG and primer 2, 5'-CTGTCCACATACTCTCCCCACAT; and for Hexb were, primer 4, 5'-GGTTTCTACAAGAGACATCATGGC and primer 5, 5'-CAATCGGTGCTTACAGGTTTCATC. Primers for detection of the PGKneo vector sequences were primer 3, 5'-CACCAAAGAAGGGAGCCGGT in the pgk promoter and primer 6, 5'-GATATTGCTGAAGAGCTTGGCGGC in neo. These primers (final concentration of 0.5 [mu]M) were added to a PCR buffer containing 20 mM Tris (pH 8.4), 50 mM KCl, 1.5 mM MgCl2, 0.2 mM dNTPs, 2.5 units Taq DNA polymerase (GIBCO) and 2 [mu]l of genomic DNA. The PCR samples (total volume, 100 [mu]l) were preheated at 95oC for 5 min. Thirty-five PCR cycles were performed composed of 20 s denaturation at 94oC, 20 s annealing at 57oC, and 1 min extension at 72oC. The products (16 [mu]l) were analyzed on a 2% agarose gel.
Total cellular RNA was prepared from various mouse tissues including testis, epididymis, lung, liver, kidney, heart, spleen and total brain and analyzed by Northern blot as described previously (17 ,18 ). The hybridization probes from the Hexa and Hexb cDNA contained the entire coding sequences. A 600 bp fragment from the neomycin-resistance gene was used to detect the neo sequence, and a synthetic oligonucleotide recognizing 18S ribosomal RNA was used to assess equivalence of loading between lanes.
Proteins were isolated from total brain and liver from Hexa and Hexb +/+ and -/- mice. Tissues were homogenized in 5 vol (wt/vol) of cold 0.025 M imidazole (pH 7.4). The homogenates were alternately frozen-thawed three times, briefly sonicated on ice and spun down at 12 000 g for 10 min at 4oC. Aliquots containing 1 OD280 of total soluble protein were applied to a chromatofocusing column at 4oC containing the Polybuffer Exchanger (PBE 94, Pharmacia Fine Chemicals) as described elsewhere (20 ). Hex activity of each fraction was determined as described (43 ), using 3 mM MUG (SIGMA) and MUGS (Toronto Research Chemicals). The fluorescence was determined in an LS 50 Luminescence Spectrometer (Perkin Elmer) with the excitation wavelength at 360 nm and the emission wavelength at 447 nm. The pH of each fraction was also determined.
The murine mAb 2D4 recognizing GA2 was obtained from ATCC (ATCC TIB 185). The mouse/human chimeric IgG1 antibody KM966 recognizing GM2 ganglioside was prepared as described (44 ). HRP-conjugated-goat antihuman IgG ([gamma]) was from BioSource International, FITC-conjugated goat F(ab')2 anti-human IgG were obtained from Southern Biotechnology Associates and HRP-conjugated rabbit antimouse IgM was from Zymed Laboratories.
Glycosphingolipids were extracted, purified and characterized as described in detail earlier (45 ,46 ). In brief, murine tissues (0.1-0.5 g wet weight) were extracted with chloroform/methanol solvents. Following saponification, the glycosphingolipids were fractionated into neutral and acid glycolipids by DEAE-Sephadex A-25 chromatography. Glycosphingolipids were separated into individual components by HPTLC on Si-60 plates (E. Merck) in solvents chloroform/acetone 1:1 v/v (pre-run) followed by chloroform/methanol/water 65:25:4 v/v (neutral glycosphingolipids) and chloroform/methanol/0.2% CaCl2 55:45:10 v/v (gangliosides). Neutral glycosphingolipids were visualized with orcinol/H2SO4 and gangliosides with resorcinol/HCl. GM2 and GA2 were identified by immune thin layer chromatography using mAbs to GM2 (KM966 5 [mu]g IgG/ml) and GA2 (2D4, supernatant 1:2 diluted), with purified bovine brain GA2 and GM2 (Alexis Corp., San Diego, CA) as standards and diaminobenzidine as chromogen. GM2 and GA2 were quantitated by fluorescence immune thin-layer chromatography using antibody KM966 (5 [mu]g IgG/ml) or 2D4 (supernatant diluted 1:2) and FITC-conjugated antihuman IgG or antimouse IgM antibody and a fluor image analyzer (FluoroImager 575, Molecular Dynamics). The quantitations were made within the linear range of calibration curves determined for authentic standards (data not shown).
Mice were anesthetized with sodium pentobarbital and tissues were perfused-fixed through the heart with 2.5% glutaraldehyde buffered in sodium cacodylate (0.1 M) containing 0.5% CaCl2 (pH 7.4). Brain, liver, kidneys and spleen were removed and placed in the same fixative for 1 h at 4oC. Following fixation and washing in 0.1 M cacodylate buffer the tissues were postfixed in ferrocyanide-reduced osmium for 1 h at 4oC, dehydrated in a graded ethanol series and embedded in Epon (47 ). Semi-thin sections (0.5 [mu]m thick) were cut and stained with toluidine blue and examined under the light microscope. Thin sections (65 nm thick) were mounted on slotted copper grids, stained with uranyl acetate and lead citrate and examined with an electron microscope.
For immunohistochemical detection of GM2 ganglioside in tissue sections, anesthetized mice were perfused through the heart with freshly made 4% paraformaldehyde. Following perfusion, brain and spinal cord were removed and further fixed in the same fixative overnight at 4oC. The samples were rinsed briefly in 0.1 M phosphate-buffered saline, dehydrated in 0.5 M sucrose solution and snap frozen in chilled isopentane. Tissues were stored at -80oC. For immunohistochemical staining, 10-15 [mu]m frozen sections were cut in a cryostat and collected on to poly-L-lysine coated slides. The sections were post-fixed in 4% paraformaldehyde for 10 min and rinsed in Tris-buffered saline (TBS). Endogenous peroxidase activity was blocked with 0.3% H2O2 in methanol for 30 min and rinsed in TBS. The slides were treated with 0.1% Triton X-100 in TBS, followed by 1.5% bovine serum albumin solution. The slides were incubated with the KM966 antibody diluted to 1/600 in TBS for 1 h at room temperature in a humidified chamber. They were rinsed, incubated with biotinylated goat antihuman IgG antibody diluted to 1/200 in TBS for 30 min, rinsed, incubated with ABC reagent (Vector Laboratories) for 30 min and developed in diaminobenzidine. The slides were counterstained with haematoxylin, dehydrated through ethanol and xylene and mounted in permount.
We thank the Canadian Genetic Diseases Network for providing access to its gene targeting core facilities and travel support to D. P. and N. W. We thank P. Hechtman and D. Mahuran, M. Fernandes and D. Trasler for advice and dicussion, K. Franklin for consultations on the neuropathology and for making plates from his forthcoming A Stereotaxic Atlas of the Mouse Brain (K. Franklin and G. Paxinos, eds, Academic Press) available prior to publication, M. Shevell for assistance with the phenotype, L. Old for helping make the ganglioside analysis possible, B. Triggs-Raine for providing Hexb genomic clones and reading the manuscript, and M. Aardse for keeping track of us. D. P. was supported by a scholarship from the Fonds de recherche en santé du Québec, N. W. from the Medical Research Council of Canada. This project was supported by grants to R. A. G. from the Canadian Genetic Diseases Network through the Medical Research Council of Canada and to J. M. T. from the Medical Research Council. The Hexb -/- mouse was developed by D. P., the Hexa -/- mouse by N. W.
1 Ledeen, R. (1985) Gangliosides of the neuron. TINS, 8, 169-174.
2 Kracun, I., Rosner, H., Cosovic, C., and Stavljenic, A. (1984) Topographical atlas of the gangliosides of the adult human brain. J. Neurochem.43, 979-989. MEDLINE Abstract
3 Svennerholm, L., Bostrom, K., Fredman, P., Jungbjer, B., Lekman, A., Mansson, J.E, and Rynmark, B.M. (1994) Gangliosides and allied glycosphingolipids in human peripheral nerve and spinal cord. Biochim. Biophys. Acta, 1214, 115-123. MEDLINE Abstract
4 Svennerholm, L. (1980) Gangliosides and synaptic transmission. In Svennerholm,L., Mandel,P., Dreyfus,H., and Urban,P.-F.(eds) Structure and Function of Gangliosides, Plenum Press, New York, p. 533.
5 Schengrund, C.L. (1990) The role(s) of gangliosides in neural differentiation and repair, a perspective. Brain Res. Bull.24, 131-141. MEDLINE Abstract
6 Gravel, R.A., Clarke, J.T.R., Kaback, M.M., Mahuran, D., Sandhoff, K., and Suzuki, K. (1995) The GM2 gangliosidoses. In Scriver,C.R., Beaudet,A.L., Sly,W.S., Valle,D. (eds) The Metabolic and Molecular Bases of Inherited Disease, McGraw-Hill, New York, pp. 2839-2879.
7 Kytzia, H.J. and Sandhoff, K. (1985) Evidence for two different active sites on human beta-hexosaminidase A. Interaction of GM2 activator protein with beta-hexosaminidase A. J. Biol. Chem.260, 7568-7572. MEDLINE Abstract
8 Conzelmann, E. and Sandhoff, K. (1978) Deficiency of a factor necessary for stimulation of hexosaminidase A-catalyzed degradation of ganglioside GM2 and glycolipid GA2. Proc. Natl Acad. Sci USA, 75, 3979-3983. MEDLINE Abstract
9 Hechtman, P., Gordon, B.A., and Ng Ying Kin, N.M. (1982) Deficiency of the hexosaminidase A activator protein in a case of GM2 gangliosidosis; variant AB. Pediatr. Res.16, 217-222. MEDLINE Abstract
10 Snyder, E.Y., Taylor, R.M., and Wolfe, J.H. (1995) Neural progenitor cell engraftment corrects lysosomal storage throughout the MPS VII mouse brain. Nature, 374, 367-370. MEDLINE Abstract
11 Cummings, J.F., Wood, P.A., Walkley, S.U., de Lahunta, A., and DeForest, M.E. (1985) GM2 gangliosidosis in a Japanese spaniel. Acta Neuropathol. (Berl), 67, 247-253.
12 Karbe, E. and Schiefer, B. (1967) Familial amaurotic idiocy in male German Shorthair Pointers. Vet. Pathol.4, 223-232.
13 Kosanke, S., Pierce, K., and Bay, W. (1978) Clinical and biochemical abnormalities in Porcine GM2-gangliosidosis. Vet. Pathol.15, 685-699. MEDLINE Abstract
14 Cork, L.C., Munnel, J.F., Lorenz, M.D., Murphy, J.V., Baker, H.J., and Rattazzi, M.C. (1977) GM2 ganglioside lysosomal storage disease in cats with beta-hexosaminidase deficiency. Science, 196, 1014-1017. MEDLINE Abstract
15 Neuwelt, E.A., Johnson, W.G., Blank, N.K., Pagel, M.A., Maslen McClure, C., McClure, M.J., and Wu, P.M. (1985) Characterization of a new model of GM2-gangliosidosis (Sandhoff's disease) in Korat cats. J. Clin. Invest.76, 482-490. MEDLINE Abstract
16 Muldoon, L.L., Neuwelt, E.A., Pagel, M.A., and Weiss, D.L. (1994) Characterization of the molecular defect in a feline model for type II GM2-gangliosidosis (Sandhoff disease). Am. J. Pathol.144, 1109-1118. MEDLINE Abstract
17 Wakamatsu, N., Benoit, G., Lamhonwah, A.M., Zhang, Z.X., Trasler, J.M., Triggs-Raine, B.L., and Gravel, R.A. (1994) Structural organization, sequence, and expression of the mouse HEXA gene encoding the alpha subunit of hexosaminidase A. Genomics, 24, 110-119. MEDLINE Abstract
18 Triggs-Raine, B.L., Benoit, G., Salo, T.J., Trasler, J.M., and Gravel, R.A. (1994) Characterization of the murine beta-hexosaminidase (HEXB) gene. Biochim. Biophys. Acta, 1227, 79-86. MEDLINE Abstract
19 Beccari, T., Orlacchio, O., and Stirling, J.L. (1988) Identification of [beta]-N-acetylhexosaminidase A in mouse tissues with the fluorigenic substrate 4-methylumbelliferyl-[beta]-N-acetylglucosamine 6-sulphate. Biochem. J.252, 617-620. MEDLINE Abstract
20 O'Dowd, B.F., Klavins, M.H., Willard, H.F., Gravel, R., Lowden, J.A., and Mahuran, D.J. (1986) Molecular heterogeneity in the infantile and juvenile forms of Sandhoff disease (O-variant GM2 gangliosidosis). J. Biol. Chem.261, 12680-12685. MEDLINE Abstract
21 Dewji, N.N., De Keyzer, D.R., and Stirling, J.L. (1986) Purification and characterization of beta-N-acetylhexosaminidase I2 from human liver. Biochem. J.234, 157-162. MEDLINE Abstract
22 Mahuran, D.J. (1990) Characterization of human placental beta-hexosaminidase I2. Proteolytic processing intermediates of hexosaminidase A. J. Biol. Chem.265, 6794-6799. MEDLINE Abstract
23 Geiger, B., Arnon, R., and Sandhoff, K. (1977) Immunochemical and biochemical investigation of hexosaminidase S. Am. J. Hum. Genet.29, 508-522. MEDLINE Abstract
24 Hawkes, R. and Gravel, C. (1991) The modular cerebellum. [Review]. Prog. Neurobiol., 36, 309-327. MEDLINE Abstract
25 Yamanaka, S., Johnson, M.D., Grinberg, A., Westphal, H., Crawley, J.N., Taniike, M, Suzuki, K., and Proia, R.L. (1994) Targeted disruption of the Hexa gene results in mice with biochemical and pathologic features of Tay-Sachs disease. Proc. Natl Acad. Sci. USA, 91, 9975-9979. MEDLINE Abstract
26 Taniike, M., Yamanaka, S., Proia, R.L., Langaman, C., Bone-Turrentine, T., and Suzuki, K. (1995) Neuropathology of mice with targeted disruption of Hexa gene, a model of Tay-Sachs disease. Acta Neuropathol.89, 296-304. MEDLINE Abstract
27 Adachi, M., Schneck, L., and Volk, B. (1974) Ultrastructural studies of eight cases of fetal Tay-Sachs disease. Lab. Invest.30, 102-112. MEDLINE Abstract
28 Walkley, S.U., Baker, H.J., and Rattazzi, M.C. (1990) Initiation and growth of ectopic neurites and meganeurites during postnatal cortical development in ganglioside storage disease. Brain Res. Dev. Brain Res.51, 167-178. MEDLINE Abstract
29 Walkley, S.U., Baker, H.J., Rattazzi, M.C., Haskins, M.E., and Wu, J.Y. (1991) Neuroaxonal dystrophy in neuronal storage disorders, evidence for major GABAergic neuron involvement. J. Neurol. Sci., 104, 1-8. MEDLINE Abstract
30 Conzelmann, E. and Sandhoff, K. (1979) Purification and characterization of an activator protein for the degradation of glycolipids GM2 and GA2 by hexosaminidase A. Hoppe-Seyler's Z. Physiol. Chem.360, 1837-1849. MEDLINE Abstract
31 Burg, J., Banerjee, A., Conzelmann, E., and Sandhoff, K. (1983) Activating proteins for ganglioside GM2 degradation by beta-hexosaminidase isoenzymes in tissue extracts from different species. Hoppe Seyler's. Z. Physiol. Chem.364, 821-829.
32 Nagai, T. and Yamada, H. (1988) Characterization of mouse liver sialidase and partial purification of the lysosomal sialidase. Chem. Pharm. Bull.36, 4008-4018. MEDLINE Abstract
33 Miyagi, T., Sagawa, J., Konno, K., Handa, S., and Tsuiki, S. (1990) Biochemical and immunological studies on two distinct ganglioside-hydrolyzing sialidases from the particulate fraction of rat brain. J. Biochem. (Tokyo), 107, 787-793.
34 Sandhoff, K., Harzer, K., Wassle, W., and Jatzkewitz, H. (1971) Enzyme alterations and lipid storage in three variants of Tay-Sachs disease. J. Neurochem.18, 2469-2489. MEDLINE Abstract
35 Riboni, L., Caminiti, A., Bassi, R., and Tettamanti, G. (1995) The degradative pathway of gangliosides GM1 and GM2 in Neuro2a cells by sialidase. J. Neurochem., 64, 451-454. MEDLINE Abstract
36 Sango, K., Yamanaka, S., Hoffmann, A., Okuda, Y., Grinberg, A., Westphal, H., McDonald, M.P., Crawley, J.N., Sandhoff, K., Suzuki, K., and Proia, R.L. (1995) Mouse models of Tay-Sachs and Sandhoff diseases differ in neurologic phenotype and ganglioside metabolism. Nature Genet.11, 170-176. MEDLINE Abstract
37 Soriano, P., Montgomery, C., Geske, R., and Bradley, A. (1991) Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell, 64, 693-702. MEDLINE Abstract
38 Beccari, T., Hoade, J., Orlacchio, A., and Stirling, J.L. (1992) Cloning and sequence analysis of a cDNA encoding the alpha-subunit of mouse beta-N-acetylhexosaminidase and comparison with the human enzyme. Biochem. J., 285, 593-596. MEDLINE Abstract
39 Bapat, B., Ethier, M., Neote, K., Mahuran, D., and Gravel, R.A. (1988) Cloning and sequence analysis of a cDNA encoding the beta-subunit of mouse beta-hexosaminidase. FEBS Lett.237, 191-195. MEDLINE Abstract
40 Nagy, A., Rossant, J., Nagy, R., Abramow-Newerly, W., and Roder, J.C. (1993) Derivation of completely cell culture-derived mice from early-passage embryonic stem cells. Proc. Natl Acad. Sci. USA, 90, 8424-8428. MEDLINE Abstract
41 Wurst, W. and Joyner, A.L. (1993) Production of targeted embryonic stem cell clones. In Joyner,A.L. (ed.), Gene Targeting: A Practical Approach, IRL Press, New York, pp. 33-61.
42 Papaioannou, V. and Johnson, R. (1993) Production of chimeras and genetically defined offspring from targeted ES cells. In Joyner,A.L. (ed.), Gene Targeting: A Practical Approach, IRL Press, New York, pp. 107-146.
43 Suzuki, K. (1987) Enzymatic diagnosis of sphingolipidoses. Methods Enzymol.138, 727-762. MEDLINE Abstract
44 Nakamura, K., Koike, M., Shitara, K., Kuwana, Y., Kiuragi, K., Igarashi, S., Hasegawa, M., and Hanai, N. (1994) Chimeric anti-ganglioside GM2 antibody with antitumor activity. Cancer Res., 54, 1511-1516. MEDLINE Abstract
45 Momoi, T. and Wiegandt, H. (1980) Separation and micro-detection of oligosaccharides of glycosphingolipids by high performance cellulose thin-layer chromatography-autoradiofluorography. Hoppe-Seyler's Z. Physiol. Chem.361, 1201-1210. MEDLINE Abstract
46 Ritter, G., Krause, W., Geyer, R., Stirm, S., and Wiegandt, H. (1987) Glycosphingolipid composition of human semen. Arch. Biochem. Biophys.257, 370-378. MEDLINE Abstract
47 Morales, C.R. and Hermo, L. (1983) Demonstration of fluid-phase endocytosis in epithelial cells of the male reproductive system by means of horseradish peroxidase-colloidal gold complex. Cell Tissue Res.230, 503-510. MEDLINE Abstract
*To whom correspondence should be addressed at: McGill University-Montreal Children's Hospital Research Institute, 2300 Tupper Street, Montreal, Canada H3H 1P3
+Present address: Institute for Human Gene Therapy, University of Pennsylvania Medical Center, Philadelphia, PA, USA
Present address: First Department of Internal Medicine, University of Tokushima, Tokushima, Japan
Present address: Howard Hughes Medical Institute, University of California at San Diego, La Jolla, CA, USA
This page is maintained by OUP admin. Last updated Thu Oct 31 15:08:29 GMT 1996. Part of the OUP Journals World Wide Web service.Copyright Oxford University Press, 1996
